Background
There is strong evidence linking inflammation and the development of pancreatic ductal adenocarcinoma. Cyclooxygenase-2 (COX-2) and COX-2-derived PGE2 are overexpressed in human and murine pancreatic ductal adenocarcinoma. Several studies have demonstrated an important role of COX-2-derived PGE2 in tumor-stroma interactions; however, the direct growth effects of prostaglandin E2 (PGE2) on pancreatic ductal adenocarcinoma cells is less well defined. Our
Conclusion
Our study provides evidence that PGE2 can inhibit directly pancreatic ductal adenocarcinoma cell growth through an EP4-mediated mechanism. Together with our gene expression and survival analysis, this observation suggests a protective role of EP4 receptors in human pancreatic ductal adenocarcinoma that expresses E-type prostaglandin receptors.
Methods
Human pancreatic ductal adenocarcinoma cell lines, Panc-1 and MIA PaCa-2, were treated with PGE2 in varying doses (0-10 μM). Effects on the phosphorylation of ERK1/2 were evaluated by Western blot. Colony formation was observed for cells treated with PGE2 for 11 days. DNA synthesis was determined by (3H)-thymidine incorporation assay. Gene expression of E-type prostaglandin (EP)2/EP4 receptors and their correlation with survival in patients with pancreatic ductal adenocarcinoma were assessed using the RNA-Seq data set from The Cancer Genome Atlas Research Network.
Results
PGE2 decreased the size and number of colonies in Panc-1 but not MIA PaCa-2 cells. In the Panc-1 cells, PGE2 activated PKA/CREB and decreased phosphorylation of ERK1/2, which was reversed by an EP4 receptor antagonist, while an EP2 receptor antagonist had no effect. In contrast, in MIA PaCa-2 cells, PGE2 had no effect on ERK1/2 phosphorylation. Treatment of both Panc-1 and MIA PaCa-2 cells with forskolin/IBMX decreased ERK1/2 phosphorylation. Finally, PGE2 decreased DNA synthesis only in Panc-1 cells, which was reversed by an EP4 receptor antagonist. In human pancreatic ductal adenocarcinoma, high EP2 and low EP4 gene expression was correlated to worse median overall survival (15.6 vs 20.8 months, log-rank P = .017).